Interview With Josh Grafton Of Strangely Interactive, Developers Of Angels Fall First

I recently had the chance to sit down with Josh Grafton of Strangely Interactive to discuss their FPS Angels Fall First. We talked a bit about the previous Angels Fall First game, the mod that Angels Fall First evolved out of, and development philosophies.

Me: Is Angels Fall First a direct evolution of Angels Fall First: Planetstorm?

Grafton: It is Planetstorm. We took that word away because it sounded very Planetside-y.

So it is pretty much exactly Planetstorm with seven more years of development?

Pretty much. For the moment it’s even got two familiar maps.

What happened to the previous Angels Fall First game, The Second Antarean War?

The guy went on to do something infinitely cooler. is where Treeform, the developer, started a new game and it’s basically The Second Antarean War but 2D, and without any of the Angels Fall First stuff, and you can design your units on the fly. It’s really, really cool.

So he went on to do something else and the rest of the team went on to make Planetstorm?

Yeah. Although he’s still our webhost. He’s a cool guy. He writes all of our internal development notices and stuff.

Why did you guys decide to move away from The Second Antarean War once he was done with it instead of continuing the project?

It was all him. He came to us originally with basically The Second Antarean War before Angels Fall First dressing. And he’s like, “Hey guys, I want some art. Can you guys help me with art?” We were like, “No, we can give you all of our art if you make your game about our spaceships.” He’s like “okay,” so we kind of married him.

We did that and then he was like, “yeah, this isn’t working out.” The Second Antarean War really didn’t work out. It was far too free. We were learning and we helped him with the design. Although he had veto on the design, we were very suggestive—well, you should do this, you should do that.

It just didn’t work out. It was what it set out to be—Homeworld with multiple persistent systems, but that was such a balance headfuck that we stopped it.

What originally inspired you guys to make Planetstorm?

This was 2003 or 2004. We were playing Battlefield 2142. We were big fans of old school games like Tie Fighter. And space was not enjoying a renaissance at that point and there were no space games. We were like “you know what, all these new shooters need space.” And we waited and we thought “yeah, of course they’ll do it. This is such an obvious thing. Combine everything.”

But games didn’t really go that way. Graphics got prettier. They became more complex in terms of stats and crafting and detail like that. But they never really went the way we wanted them to go, which is bigger, more complicated gameplay.

ArmA was a good substitute. We played a lot of ArmA. But we couldn’t get any space mods for it or anything like that.

So we just wanted bigger games. We wanted to combine everything. We were those idiot 14-year olds who went, “they should make a game with everything in it!” And we believed that so much that we made a game with everything in it. We found out why people don’t make games with everything in it. Because it’s damned difficult to make a game with everything in it that works worth a damn.

You can make a game with everything in it—that’s ArmA. Have you seen the terrible consequences of giving the player total freedom? Here is a 10 square kilometer map. I decree that you will never ever see another player. That’s where DayZ came from.

How have you guys countered that in Angels Fall First?

We do not claim to be innovative with any of our features. We just kind of combined. And what was really great about Battlefield was the focus of the fights. You were always in a firefight. There were some quieter moments, but you were never far from having a good time and having some moments.

We took that as our cue to make things smaller than ArmA. But where Battlefield fell down for us was that it was still “spawn, die, spawn, die, spawn, die” but on a bigger scale. Our antithesis is Call of Duty. We didn’t want Call of Duty.

So we took the focus of Battlefield and the freedom of ArmA and tried to make a compromise. And we thought that that would work for our ground-based gameplay and also for the fleet combat.

So basically, you give people a lot to fight over and hope that they’re going to go for it?

Yeah, and the real question was “how much do we give people before it becomes meaningless?” Before it becomes a big mess of nothing and role-playing game that only serious people can get into. It’s very familiar, I hope.

I was surprised that you went for both the Newtonian physics model and a more flight assisted model for spacecraft rather than just a flight assisted model.

Oh, well you know we’re Battlestar fans as well as Star Wars fans, so we wanted to do those backflips—those crazy backflips.

The concept of having two flight modes was something that I fought for. The implementation has been handled and molded by other people. So many other people that it’s quite rounded and ergonomic just by dint of having been through so many hands. And also the code was really messy. We’re cleaning that up.

There are a lot more maps than I imagined based on the trailer.

This is nothing. We’ve got so many on the backburner. Can you imagine what happens after seven years of development with bored artists? The coders needed that time to learn how to make a game and I needed that time to learn how to make artwork efficiently. But apart from that, we’ve just been building for seven years. So we’ve got a lot on the backburner that we’re going to polish up and drizzle out.

Are the AI designed to be a replacement for people or just a stand-in?

Angels Fall First was designed to fail. It was designed to have a launch with no players. We wanted an offline experience that was comparable to an online experience.

We also wanted support for a partially player-populated online experience. So the bots are there to replace players in any role. The one thing we didn’t do is APCs, which proved to be a real nightmare. Getting bots to group together then drive an APC to another place was a horrendous nightmare. We just cut that feature from the game.

They know how to do everything that you’re supposed to do in the game. They run away. That’s the best thing that you can program an AI to do.

So they’re competent, then?

They know what they’re supposed to do. [laughs]

Then you guys set out from the beginning to make sure there was a competent AI experience? A lot of games recently seem to be setting out to abolish bots.

Yeah, we didn’t like that. Because not all of us are amazing players and want to get our asses kicked 20 times in a row upon first joining the game?

There is a good reason for it. The seven years isn’t there for nothing. That’s where the time went. AI. Definitely AI. Making bots and designing the game around them.

Because it’s not just the bots. All of these maps have navigation mesh which is completely different from the visual mesh that you see, and that’s for bots to find their way around. That was a considerable time sink.

Also the collision of everything. Bots will get stuck on collision and they won’t know that. So everything has to be designed around the bots. And that was the biggest limitation on what we’d done.

How many people worked on Angels Fall First? I saw that there were you and one other member listed on your press kit.

[laughs] Yeah, and somebody went “it’s a two-man job” and everybody went “what the hell?”

Yeah. Angels Fall First has been a project. It was a mod team and people come and go a lot. And when there’s been money and motivation—for example. We did the Make Something Unreal contest in 2009 and there was a prospect of money—naturally, we had a lot of talented guys go “hey, I could do wonders for your team.” And then, when there hasn’t been money involved, in the quiet times, it’s just been our diehard guys.

The lineup at the moment hasn’t changed for a few years. Our tech lead, our lead programmer, myself, and our network programmer. Four or five guys. Our community manager, as well, has been with us for a long time.

But the other guys still hang out occasionally. It’s like a hippie commune of game development. Angels Fall First was developed over IRC. I have never spoken to 90% of the team.

If you could estimate how many people have worked on it over the years…

Oh, 30 or 40 over the years. Never more than 15 at a time. Rarely the same lineup twice. Some people’s work has been replaced and some people’s work remains.

You had said earlier about not wanting to have your ass handed to you repeatedly the first time you play and that reminded me of my first time playing Planetside 2. You’re designed to be part of a meat grinder.

Yeah, we call that “the Last Starfighter experience.” We took that for space but it works for ground as well. What’s the most fun thing to do in a space game? Wade through hordes of mindless enemy.

So it’s more about the space game. Because ground fighting, you can get away with multiplayer only, but no one wants to be the underdog in space. Fighting player vs player in space. It’s very fast, very twitchy, even with all the arcade-y styles that we’ve put on.

So filling space with dumb bots was the best thing we could do for the space game. And it turned out that that was a really fun thing to do on the ground, as well.

I put the game on in the background sometimes. Ambient...battle.

Since it is an evolution of Planetstorm, are you still using UDK?

As near as dammit. The full version, we’ve been making source changes now. We’ve got a nice relationship with Epic. We’ve got the full version. So it’s not technically UDK anymore, although it is not far from it.

What kinds of things are you looking to do with future maps? I see you’ve got a few space maps and a few ground maps, and they’re all massive.

Yeah. We want to do more with hooking them up together. People want us to do space to ground and proper planetary assault. And we keep explaining how insane that is and why that would never ever happen. But they keep on asking and we keep on thinking about it.

Our main defense is that space and ground are 80km apart in most cases and that’s just not fun. You don’t want to do 80km of anything.

We reckon it could be done with a certain amount of suspension of disbelief. I could imagine a scenario where you’re trying to get your dropships into an atmospheric entry window zone of some kind and, at that point, they cutscene down to the planet and are deployed like regular modules. That’s something that we could do without any additional functionality from our game.

Our standard space game became clearing space defenses then boarding a station. That seems to work out quite well. So we’ll do that a few times and see if we can improve on that. Become bigger and better with that.

We’ve got a couple of maps on the backburner with multiple stations to board at the same time. One which is—have you ever seen Silent Running, the 70s sci-fi movie? It’s like this giant, conservatory, hydroponic ship. There’s like this garden inside and it’s got these big domed windows. We’ve got a couple like that and there’s various biospheres. There’s a fungal jungle biosphere and there’s an arid desert biosphere and an ice biosphere and that kind of thing. They’re kind of cool. You’ve got to steal all the bio data and get back out into space again.

But we run into this problem of overextending ourselves even though we are smaller than ArmA and all those other crazy free games. Populating those two space stations whilst there’s a space battle outside? On 64 players, it’s still a little bit thin.

With humans, it might get better though. A lot of our maps have completely changed now that we’re filling them with humans. Some of them, we guessed right. Some of them, we didn’t guess so well. A lot of spawn-camping in space. We’ve dealt with that in the next patch. We’re going to put in a few new jump points.

So you are facing issues with scale vs how many people are playing?

Yeah. We couldn’t afford to get professional testing. We didn’t have the attention to even fill a server. Now we’re learning our lessons about public scale. And a lot of people could complain that we’re putting the cost on the user by charging for it now but we still think you’re getting a really good value out of Angels Fall First.

How important is it that Early Access sales go well?

It’s not important for us. Maybe for the continued popularity and success of the game. In the short term, for the playerbase, we are worried about it. We want to do our best to fill servers so that people have a good experience. But we’re not actually in it for the money. We’re in it to make this game. And this is why we did it the slow way. This is why we’re doing it part time.

So then you’ve already worked on the game this long part-time and you’re not concerned with how it sells?

Not initially for our benefit, no. We are concerned to provide a good experience for the players that are supporting us. So we will be promoting it and we will be doing events and we will be trying to fill the servers.

How does the points system for your loadout work?

You earn points from one of three channels: combat, command, or support. Different weapons require different combinations of points cost. And when you have the amount of points that weapon requires, you can equip it.

Do you earn them on a per-life basis?

No, they should be permanent. Such is the nature of our development that the guys will often make little quirks and tweaks and not tell anyone for a while. One of the things I found out is that apparently, it will give you a few extra points within matches. I haven’t seen this happen.

So you earn the points at the end of matches, then?

Yeah. At the end of the match, your points get updated. Within a match it adds a few more points, but I am hazy on that mechanic at the moment.

The points system is generally designed to be permanent and right now we’re looking at dividing up online and offline matches so that you don’t earn points in the offline matches. At the moment you kind of earn points, but they’re not saved and that’s kind of confusing some people. And it should be that you don’t earn points in offline matches. And maybe we’ll put some kind of ranked registration on the server so that you can switch ranking on and off as a server option.

In offline matches will you have unlimited points or will you not earn points at all?

For now, we’ll just turn the points earning off. As for raising the amount of points offline, that could be something to consider but it’s always so galling when you go back online and you don’t have the stuff.

You can have quite a lot of stuff at most times, due to the nature of how it’s dished out. It’s never that a weapon is behind a hard gate. It’s if you can scrape together the points and sacrifice enough other things to have it.

Where do you resupply when you run out?

You’ve got a knife. You can use that to go hunting for new guns. If someone’s dropped an ammo pack you can resupply there. There have been a lot of requests to have a resupply in the dropship, which we pretty much agree with. So we might implement that soon.

It can be quite bit brutal. We’ve also taken a lot of flack about our high TTK—time to kill. People say it feels a bit bullet spongey. Well, sometimes you are that bullet sponge and I hope you’re glad at those moments that it’s a high-TTK game. We wanted people to be able to survive the first attack and to step back and see what they were doing wrong.

Especially with some of the large spaces that we have. We’ve necessarily got a long run from spawn and that is not fun if it happens a lot.

Is damage localized?

No, it’s not localized to your armor on your characters. That was a bit overkill. We thought about all these complicated and cool systems for the localization of damage, but this is quite a fast-paced game and nobody really notices because you’re dead.

I think what we have to work on is perceived TTK. We could do a lot with our netcode. That’s affecting hitboxing right now so people are uncertain if it’s their fault. We need to make them certain that it’s their fault.

We also have a little bit of money now. We’ve already started hiring people actually.

So your team is going to expand in the near future?

We’ve hired some contractors to improve maps in the short term but we’d like to keep them.

We should have those hover tanks able to run you over. We used to have prone. We took it out because it was a lot of hassle. Things like the collision hitboxes wouldn’t align properly and horrible, horrible problems so we just ditched it.

I kind of miss it. Especially when I think that we could have made it so that when you were prone, the hover tanks would go straight over you.

Despite how much Angels Fall First has it is but the tip of our imagination iceberg.

Are you planning to expand the weapon set more?

Yeah, absolutely. We’ve had mixed comments about them. “They’re just a bunch of bullet throwers.” Yeah, okay, we can do some newer crazier ones, but I like bullet throwers. I like the mystery in having a lot of similar guns. Is this one better? I like the way this one feels. I like the way this one looks. It’s nuance.

There are guys on the team that wanted to just go “no, one lead thrower, one plasma gun, one laser gun.” I find this terribly black and white. This doesn’t exist in the real world. So there are a few similar guns. And there will be more similar guns and more crazy guns, as well. Mods, guns, they’re something we can do without much expenditure. It’s something that the permanent guys can make in our spare time.

Are you planning on allowing modding?

We would have to renegotiate our contract with Epic, I think. It requires some special dispensations that we don’t currently have. It’s not out of the question. What might be out of the question is letting people in to that degree. It’s not working as well as it could. So if we do let people in, and allow people make mods we’d want it to be a lot tighter before that happened. So not in the foreseeable future, no.

You estimated that you were going to have about a year in Early Access. Will it be mostly polishing or is there still more that you want to add?

There’s a lot of stuff we want to add. We’ll think of more stuff, and yes, polishing, as well.

Do you foresee that extending at all with the whole “think of more stuff” part?

No, I don’t think we’ll spend much more than a year in Early Access. We’re all getting old. We want to have babies and stuff. And that’s not very conducive to game development.

Once it’s out of Early Access is it still going to be supported regularly?

Once it’s out of Early Access we’ll start on Angels Fall First 2. Much to the dismay of our wives and babies.

So you are planning to keep going from here?

Yeah, how could we not? Look how cool this thing is.

When moving to a new game, what do you think you’d change from the get go?

What the game is built on. We will move to Unreal 4 if we do a sequel. We will move to Unreal 4, we will rewrite things from scratch, it will be liberating.

So the engine is a major part of what might be changed?

Yeah, because there is no porting apart from game assets. We couldn’t port the maps. We couldn’t the code. So things would be done from scratch. Completely new tools, completely new system. I think, because of the nature of our alliance in Angels Fall First, the nature of the question would be “what does the team want to do?”

These are my dreams. I would like to carry on in Unreal 4. I know a couple of guys want to do this, too. But like I said, we’re getting old. The lineup may change again. We may decide to do something different. We may decide to stop. I would be sad. I don’t think I’ll ever stop. I think I’ll just keep going. But who knows.

So there are no cockpits?

Nope. Economy game. No cockpits—yet.

Does that mean it’s under consideration?

Yeah, a lot of people have asked for it. But if you look at the spaceship there is no cockpit. Why would you have a glass window on the front of your space fighter when there are perfectly good cameras. However, Titanfall. Yeah, Titanfall. We like that. That could work.

Yeah, cockpits. It would be cool, but it’s visual fluff. Also, right now, our control method is not centered around locking of view. It’s something that we’ve had lots of fights about. And we’ll probably continue to have lots of fights about.

Is the entire battleship modeled on the inside to scale?

No, not to scale. Space is 1/20 scale. Right now, you are flying around in a 1/20 model. If you can get inside the ship maps—and that has happened on occasion, and it is hilarious—you are flying a tiny fighter craft. And you can fly around people’s heads like “ah, I’m a little tiny mosquito.”

Yeah, so that’s how we did it. Space is very, very small and outside space, in some unspecified area, there are the internal maps.

Inside has a full interior that matches up with the outside. But it’s not to scale. We do all sorts of trickery.

I noticed that what you see out of windows is what is actually around the your ships, though.

Oh, not only that but the HUD works on them. You can still select targets out of the window. You can stand on the bridge of your battleship and command from there by pointing out the window and going “all weapons, open fire.”

What made you decide to go from a mod to a full game?

Good people cost money. We could keep trolling university undergrads. We could keep trolling mod forums and stuff, but the scene’s not as strong as it was because games are more locked down than they used to be.

So the mod was a proof of concept?

Yeah, pretty much.

Is it a progression of development that you would follow in the future? Starting as a mod and working towards a full game.

Would we go back to modding? No, not necessarily. It was a means to an end, getting people together to work on something. If we can do that by paying them then that’s even better.


Angels Fall First is currently in Early Access and can be purchased for $17.99.

I've been playing MMOs since back in the day when my only option was to play Clan Lord on the family Mac. Since then, I've played too many MMOs to count. I generally play niche, sometimes even bizarre, MMOs and I've probably logged the most hours in Linkrealms prior to its current iteration. Currently bouncing between a few games.